Serverless

DeepInfra

Open-model inference from $0.02 per million tokens.

Fireworks AI

Production inference platform for open-weights LLMs

Hyperbolic

GPU marketplace and pay-per-token inference for 25+ open AI models

Inngest

Durable execution for backend and AI workflows, no queues to run.

LanceDB

Embedded, serverless vector DB for multimodal AI, no server to run

Modal

Serverless GPU compute for AI workloads

Parasail

Brokered GPU inference: pay per token, skip the long-term contract

Pinecone

Serverless vector database with pay-per-use pricing

Turbopuffer

Vector + BM25 search on S3 — cheap when data goes cold